Food Microbiology 78 (2019) 143–154

Contents lists available at ScienceDirect

Food Microbiology

journal homepage: www.elsevier.com/locate/fm

The role of the membrane lipid composition in the oxidative stress tolerance T of different wine yeasts ∗ Jennifer Vázqueza, Karlheinz Grillitschc, Günther Daumb, Albert Masa, Gemma Beltrana, , María Jesús Torijaa a Wine Biotechnology Research Group, Dept Biochemistry and Biotechnology, Faculty of Oenology, University Rovira i Virgili, 43007, Tarragona, Catalonia, Spain b Institute of Biochemistry, Graz University of Technology, 8010, Graz, Austria c Austria Centre of Industrial Biotechnology, 8010, Graz, Austria

ARTICLE INFO ABSTRACT

Keywords: Oxidative stress is a common stress in yeasts during the stages of the winemaking process in which aerobic Saccharomyces cerevisiae growth occurs, and it can modify the cellular lipid composition. The aim of this study was to evaluate the Non-Saccharomyces oxidative stress tolerance of two non-conventional yeasts (Torulaspora delbrueckii and Metschnikowia pulcherrima) Phospholipids compared to Saccharomyces cerevisiae. Therefore, their resistance against H2O2, the ROS production and the Fatty acids cellular lipid composition were assessed. The results showed that the non-Saccharomyces yeasts used in this study exhibited higher resistance to H2O2 stress and lower ROS accumulation than Saccharomyces. Regarding the cellular lipid composition, the two non-Saccharomyces species studied here displayed a high percentage of polyunsaturated fatty acids, which resulted in more fluid membranes. This result could indicate that these yeasts have been evolutionarily adapted to have better resistance against the oxidative stress. Furthermore, under external oxidative stress, non-Saccharomyces yeasts were better able to adapt their lipid composition as a defense mechanism by decreasing their percentage of polyunsaturated fatty acids and and increasing their monounsaturated fatty acids.

1. Introduction enzymatic responses, and finally, the adapted cells can resume pro- liferation (Herrero et al., 2008; Jamieson, 1998; Moradas-Ferreira and Our understanding of the response and adaptation of yeasts to ex- Costa, 2000). ternal environmental changes is very important within the biotechno- Biological membranes are primarily made of proteins and phos- logical, pharmaceutical, food and beverage industries. Changes in the pholipids, and they form the first barrier that separates yeast cells and temperature, pH and osmotic pressure, nutrient starvation, ethanol their organelle compartments from their external environment. Fatty toxicity, prolonged anaerobiosis, exposure to chemical preservatives acids (FAs), both free and as part of complex lipids, play a number of and oxidative stress are the primary causes for the decrease in yeast key roles in metabolism. They can be incorporated into phospholipids viability and vitality in industrial processes (Walker and Dijck, 2006; (PLs), which are considered as primary structural elements of biological Gibson et al., 2007). membranes and sphingolipids, or they can serve as an energy reservoir Oxidative stress is the result of an imbalance between the presence in the form of triacylglycerols and steryl esters (Klug and Daum, 2014). of reactive oxygen species (ROS) and the capacity of cells to detoxify Another important and essential group of lipids for maintaining the these reactive intermediates of molecular oxygen, or to repair the re- membrane integrity is the sterols, and is the primary in sulting damage. Disturbances in the normal redox state of cells can yeast (Daum et al., 1998; Klug and Daum, 2014). Membrane dysfunc- damage all of their components, including lipids, carbohydrates, pro- tion can be associated with a loss of viability (Avery, 2011). Excessive teins and nucleic acids, and they may even induce programmed cell ROS production can overwhelm the detoxifying mechanism and initiate death (Costa and Moradas-Ferreira, 2001; Gibson et al., 2008; Moradas- changes in the lipid layers composition, resulting in a lipid peroxidation Ferreira et al., 1996). Under normal physiological conditions, yeasts are process, in which unsaturated lipids are converted into polar lipid hy- able to effectively defend themselves against the direct consequences of droperoxides. PLs are particularly susceptible to oxidative damage stress exposure and damage by immediate cellular enzymatic and non- mediated by ROS due to their content of polyunsaturated FAs (PUFAs),

∗ Corresponding author. Dept. Bioquímica i Biotecnologia, Facultat d’Enologia, Universitat Rovira i Virgili, c/ Marcel·lí Domingo no 1, 43007 Tarragona, Spain. E-mail address: [email protected] (G. Beltran). https://doi.org/10.1016/j.fm.2018.10.001 Received 23 November 2017; Received in revised form 16 April 2018; Accepted 5 October 2018 Available online 12 October 2018 0740-0020/ © 2018 Elsevier Ltd. All rights reserved. J. Vázquez et al. Food Microbiology 78 (2019) 143–154 which are more sensitive to peroxidation than monounsaturated FAs deposited in the Spanish Type Culture Collection (CECT) as CECT (MUFAs) (Ayala et al., 2014; Howlett and Avery, 1997). Extensive lipid 13135 and CECT 13131, respectively. peroxidation has been correlated with membrane disintegration and The commercial strains were in active dry yeast form and were re- cell death. However, lethal consequences on membranes are not sys- hydrated according to the manufacturer's instructions. For all experi- tematically observed because yeasts are able to sense and adapt to ments, precultures for biomass propagation were prepared in YPD li- environmental changes by modifying the membrane fluidity and phase quid medium (2% (w/v) glucose, 2% (w/v) peptone and 1% (w/v) transitions and by activating the cellular control of the chemical yeast extract (Panreac, Barcelona, Spain)) and incubated for 24 h at membrane composition. These changes in lipid composition are used by 28 °C with orbital shaking (120 rpm). yeast as a defense mechanism, and they are important for conferring resistance to oxidative stress. (Beney and Gervais, 2001; Los and 2.2. Effect of hydrogen peroxide on yeast growth Murata, 2004). Yeast species, and even different strains of the same species, can Yeast cells were pre-cultured for 24 h and then inoculated into YPD exhibit variations in their membrane lipid composition (Hunter and broth (25 mL) to obtain an initial population of 5 × 105 cells/mL. After Rose, 1972). In fact, yeast membranes are structurally and functionally 6 h (early exponential phase), sublethal oxidative stress was induced in dependent on the growth conditions, e.g., Saccharomyces cerevisiae is each strain by adding 2 mM H2O2 to the yeast culture. Yeast growth was auxotrophic for oleic acid and ergosterol under strict anaerobic condi- followed in both conditions (control and stressed cells) by measuring tions (Walker and Dijck, 2006). Thus, the lipid composition should not the optical density at 600 nm (OD600) every 30 min for 24 h, using a be considered a fixed and static characteristic of a single yeast strain microplate reader (Omega Polarstar, BMG Labtech Gmbh, Ortenberg, (Beltran et al., 2008; Hunter and Rose, 1972; Torija et al., 2003). Germany). Microplate wells were filled with 250 μL of inoculated S. cerevisiae is the primary yeast species involved in wine fermen- media. A control well containing medium without inoculum was used tation (Ribereau-Gayon, 1985; Fleet and Heard, 1993); however, many to determine the background signal. Measurements were taken every other yeast species can participate in different stages of the process 30 min after pre-shaking the microplate for 30 s at 500 rpm. All assays (Beltran et al., 2002). Currently, non-Saccharomyces yeasts are used to were performed in triplicate. produce final products with improved organoleptic characteristics

(González-Royo et al., 2015; Jolly et al., 2014). In general, these yeasts 2.3. Resistance to hydrogen peroxide (H2O2) are not able to complete the alcoholic fermentation, several studies have demonstrated that some non-Saccharomyces yeasts used with se- Yeast resistance to H2O2 was assessed using the agar diffusion quential inoculation techniques, can positively contribute to the aroma method (Bauer et al., 1966; Acar, 1980). Approximately 5 × 106 cells profile, sensory complexity and color stability of the resulting product were seeded with glass beads on YPD plates, and 6 mm blank disks were (Fleet, 2008; González-Royo et al., 2015; Mas et al., 2016; Pretorius, impregnated with 10 μL of 30% (v/v), 15% (v/v), 3% (v/v) or 0.3% (v/

2000). Thus, non-Saccharomyces species can influence the organoleptic v) H2O2 (Perdrogen™, Sigma-Aldrich, MO, USA) and placed on the agar properties of wines, increasing the volatile compounds or secondary surface after drying. One disk impregnated with 10 μL of H2O was used metabolites such as glycerol, aromatic alcohols, esters and acetates as the negative control. After 48 h of incubation at 28 °C, the diameter (Belda et al., 2017; Jolly et al., 2014; Romano et al., 2003). For in- of the inhibition haloes, including the disk, was measured with a ruler stance, Torulaspora delbrueckii has been proposed to reduce the volatile and photographed using a ProtoColHr automatic colony counter (Mi- acidity produced by Saccharomyces (Bely et al., 2008), whereas crobiology International, Frederick, USA). The means of three biolo- Metschnikowia pulcherrima is recommended for the release of some vo- gical replicates were calculated. latile thiols and terpenes in white wines, increasing the aromatic in- tensity (Belda et al., 2017). Nevertheless, despite the importance of 2.4. Determination of reactive oxygen species (ROS) these yeasts, there is still a lack of knowledge about non-Saccharomyces species compared with S. cerevisiae. Therefore, studies on the effect of The effect of H2O2 on the intracellular ROS concentration was oxidative stress on non-Saccharomyces yeasts are required, not only for evaluated in the six yeast strains. Yeast cells were inoculated into 50 mL the investigating their cellular physiology but also to acquire a better of YPD broth (5 × 105 cells/mL) and grown for 6 h (early exponential understanding of the adaptations of non-conventional yeasts in re- phase) at 28 °C with orbital shaking at 120 rpm. The cells were then sponse to the changes imposed by oxidative stress. exposed to different concentrations of2 H O2 (from 2 mM to 1000 mM) The goal of this study was to compare the effects of oxidative stress for 1 h, and the ROS were determined and compared to the control between S. cerevisiae and two species of non-Saccharomyces (T. del- (sample without exposure to H2O2). Three biological replicates were set brueckii and M. pulcherrima) on the yeast composition. To accomplish up for each condition. Determination of ROS was performed according this goal, we evaluated the H2O2 resistance, intracellular ROS produc- to the method described by Vázquez et al. (2017) using dihy- tion and the lipid composition (FAs, PLs and sterols) in these species drorhodamine 123 (DHR 123; Sigma-Aldrich) as an ROS indicator. DHR before and after oxidative stress exposure via H2O2. 123 is an uncharged and non-fluorescent compound that can passively diffuse across membranes being oxidized to cationic rhodamine 123, 2. Materials and methods exhibiting green fluorescence that can be measured by flow cytometry. The mean fluorescence index (MFI) was calculated according to 2.1. Yeast strains and growth conditions Boettiger et al. (2001): [(geometric mean of the positive fluorescence) – (geometric mean of the control)]/(geometric mean of the control). The yeast strains used in this study were as follows: two strains of S. Additionally, cell viability was evaluated in cells previously exposed to cerevisiae (the laboratory strain BY4742, (EUROSCARF collection, 100 and 1000 mM of H2O2 using LIVE/DEAD™ BacLight Viability kit ® Frankfurt, Germany) and a commercial wine strain (QA23 )), two (Molecular Probes, Eugene, OR, USA). Briefly, 1 mL of sample was ® strains of Torulaspora delbrueckii (BIODIVA (TdB) and Tdp) and two stained with 1 μL propidium iodide (PI)/SYTO 9 (50:50) during 15 min ® strains of Metschnikowia pulcherrima (FLAVIA (MpF) and Mpp). in darkness, each sample was washed with 1 mL of PBS to eliminate the Commercial Saccharomyces and non-Saccharomyces wine strains QA23, excess dye and analyzed immediately in the flow cytometer. FLAVIA and BIODIVA were provided by Lallemand S.A. (Montreal, Canada), and the other two non-Saccharomyces strains (Tdp and Mpp) 2.5. Experimental conditions for lipid analysis were isolated from natural musts that were taken from the Priorat Appellation of Origin (Catalonia, Spain) (Padilla et al., 2017, 2016) and Cells from each of the six yeast strains were inoculated into 450 mL

144 J. Vázquez et al. Food Microbiology 78 (2019) 143–154

Fig. 1. Effect of H2O2 (2 mM) on the cell growth of different yeast strains, compared with their control condition (without oxidative stress). Arrows indicate cell exposition to oxidative stress at 6 h. Saccharomyces cerevisiae: (A) BY4742 and (B) QA23 strains, Torulaspora delbrueckii: (C) TdB and (D) Tdp strains and Metschnikowia pulcherrima: (E) MpF and (F) Mpp strains. Growth curves of all control (G) and stressed cells (H). of YPD broth to obtain an initial population of 5 × 105 cells/mL and 10 min. Aliquots of homogenates were precipitated with 10% (v/v) of grown at 30 °C with orbital shaking at 130 rpm. After 6 h (early ex- trichloroacetic acid to quantify the protein amount with the Folin ponential phase), sublethal oxidative stress was induced in each strain phenol reagent (Lowry et al., 1951). Total lipids were extracted from by adding 2 mM H2O2 to the yeast culture. Cells were harvested for cell yeast homogenates corresponding to 1 mg, 3 mg or 0.5 mg of total subsequent lipid analysis (OD600 ∼ 10) at 6 h (before stress) and 18 h cell protein, in order to analyze FA, PL or sterol assays, respectively, after the stress exposure (thus, 24 h from the beginning of the experi- according to Folch et al. (1957). ment), in order to allow the cells to respond/adapt to this stress. Two biological replicates were set up for each strain. 2.6.2. Fatty acids The cell FA composition was analyzed by gas liquid chromato- 2.6. Lipid analysis graphy (GLC) according to Rußmayer et al. (2015). In brief, the total FAs from lipid extracts (1 mg of total cell protein) were converted to 2.6.1. Cell homogenates, protein quantification and lipid extraction methyl esters by methanolysis with sulfuric acid (2.5% in methanol (v/ Homogenates of the yeast cells were obtained using glass beads and v)) and heating at 80 °C for 90 min. These FA methyl esters were then ® a Disruptor Genie (Scientific Industries, Inc., NY, USA) at 4 °Cfor extracted twice with light petroleum and water (3:1; v/v) by shaking on

145 J. Vázquez et al. Food Microbiology 78 (2019) 143–154

Fig. 2. Resistance to H2O2 (10 μL from H2O2 3% (v/ v)) by disk diffusion method from six yeast strains grown on YPD plates over 48 h, expressed as the mean size of inhibition haloes (cm) and ± SD of n = 3. Saccharomyces cerevisiae: (A) BY4742 and (B) QA23 strains. Torulaspora delbrueckii: (C) TdB and (D) Tdp strains. Metschnikowia pulcherrima: (E) MpF and (F) Mpp strains. Different letters in superscripts (a to e) indicate values significantly different between strains by Tukey's post-test (P < 0.05).

® a Vibrax orbital shaker (IKA, Staufen, Germany) for 30 min, and se- of oxidative stress on cell growth differed depending on the yeast strain, parated by GLC on a Hewlett-Packard 6890 gas-chromatograph (Agilent being the growth of both S. cerevisiae strains the most affected (Fig. 1). Technologies, CA, USA) using an HP-INNOWax capillary column The laboratory S. cerevisiae strain (BY4742) showed a strong inhibition (15 m × 0.25 mm x 0.50 μm film thickness) with helium as a carrier of its growth after the stress, indicating that this strain was unable to gas. Finally, the FAs were identified by comparing with a commercial adapt itself to the new stress conditions (Fig. 1A). On the other hand, all FA methyl ester standard mix (NuCheck, Inc., MN, USA) and quantified the wine yeast strains were able to grow and achieved similar OD600 using pentadecanoic acid (C15:0, Sigma-Aldrich) as an internal stan- values at 24 h than their control conditions, although with a growth dard. Two analytical replicates were used for each biological replicate. delay just after the stress exposure (Fig. 1B–F), specially S. cerevisiae QA23 (Fig. 1B). 2.6.3. Phospholipids Furthermore, all strains were plated on YPD medium, and the in- The PLs were separated by two-dimensional thin layer chromato- hibition haloes around the disks that had been previously soaked with graphy (TLC) on Silica Gel 60 plates (Merck) using chloroform: me- 3% (v/v) H2O2 were measured. The inhibition haloes for the S. cerevi- thanol: ammonia solution (25%) (65:35:5; per vol.) as the first di- siae strains (BY4742 and QA23) were 2.90 ± 0.19 cm and mension solvent and chloroform: : methanol: acetic acid: water 1.80 ± 0.12 cm, respectively (Fig. 2A and B). By contrast, the size of (50:20:10:10:5; per vol.) as the second dimension solvent (Athenstaedt the inhibition haloes was significantly smaller for all the non-Sacchar- et al., 1999). Individual PLs were visualized on TLC plates by staining omyces strains. The M. pulcherrima strains, and especially Mpp, had the with iodine vapor and then scraping the spots off the plate, which were highest resistance against 3% (v/v) of H2O2 (Fig. 2E and F; quantified by measuring the amount of phosphate (Broekhuyse, 1968). 1.2 ± 0.05 cm (MpF) and 0.75 ± 0.01 cm (Mpp)). Both T. delbrueckii The phosphate quantity was calculated as a relative amount of the total strains exhibited similar inhibition haloes, with an intermediate size phosphate (%), which was estimated as the sum of all PL spots. Two between the S. cerevisiae and M. pulcherrima strains (TdF, 1.40 ± 0.17; analytical replicates were taken for each biological replicate. Tdp, 1.40 ± 0.02; Fig. 2C and D). At lower concentrations of H2O2 (0.3%), only BY4742 showed a 2.6.4. Sterols small inhibition halo, while exposure to higher concentrations of H2O2 The individual sterol composition was determined by gas-liquid (15% and 30%) resulted in an increase in the sizes of inhibition haloes chromatography-mass spectrometry (GC-MS) after the alkaline hydro- for all the strains (Fig. S1). As with the 3% (v/v) H2O2, the S. cerevisiae lysis of the lipid extracts (0.5 mg of total cell protein) (Quail and Kelly, strains were the most affected yeasts by high concentrations of this 1996). GLC-MS was performed on a Hewlett-Packard 5690 Gas Chro- oxidant. matograph equipped with an HP 5972 mass selective detector using a capillary column (HP 5-MS; 30 m × 0.25 mm i.d. × 0.25 μm film 3.2. Determination of reactive oxygen species thickness). The injection was set at 270 °C using helium as the carrier gas with a constant flow rate set to 0.9 mLmin−1. To identify the mass For all the yeast species, the intracellular ROS levels were measured fragmentation pattern of each sterol, a solution was used as with and without H2O2 stress at the early exponential phase. Under an internal standard. The determinations were performed in duplicate. these stress conditions, S. cerevisiae strains accumulated higher amounts Free sterols and steryl esters were quantified from the homogenates. of ROS than non-Saccharomyces species (Fig. 3). BY4742 was the least

The sum of total sterols includes squalene, ergosterol precursor, and the H2O2-resistant strain (Figs. 1A and 2A), and it showed the highest levels sterol intermediates (, 4-methylzymosterol, fecosterol, 14- of ROS (Fig. 3B) followed by QA23 (Fig. 3A) and the non-Saccharomyces methylfecosterol, , and ergosterol). strains (Fig. 3C–F), with M. pulcherrima Mpp having the lowest levels of endogenous ROS (Fig. 3F). Exposure to increasing concentrations of 2.7. Data analysis H2O2, from 50 to 1000 mM, resulted in an increase in ROS accumula- tion for all yeast species (Fig. S2 A-D). Both S. cerevisiae strains showed The data were subjected to a one-way analysis of variance (ANOVA) the maximal fluorescence intensity at 50 mM of H2O2, whereas the non- and Tukey's post-hoc test to evaluate the effect of each treatment. The Saccharomyces strains reached this maximum at 500 mM. After that, the results were considered statistically significant at a p-values less than fluorescence intensity, or ROS levels, declined in all strains (Fig. S2 A- 0.05 (IBM SPSS Inc, XLSTAT Software). A Principal Component D), coinciding with the decrease on its cell viability for each strain (Fig. Analysis (PCA) was performed to visualize a 2D plot of the first two S2 E). principal components (PCs) and heatmap of relative changes in lipid composition using XLSTAT Software. 3.3. Lipid composition before and after stress exposure

3. Results First, the FA, PL and sterol compositions in the six strains in this study were evaluated after 6 h of growth in a rich medium to study the 3.1. Yeast growth and resistance to hydrogen peroxide differences in lipid composition between the three species used here(S. cerevisiae, T. delbrueckii and M. pulcherrima). The cells were then sub-

The six yeast strains were grown in YPD medium, and oxidative jected to oxidative stress (2 mM of H2O2), and the lipid composition of stress (H2O2 2 mM) was applied at early exponential phase. The effect these six strains was analyzed after 18 h to determine how the different

146 J. Vázquez et al. Food Microbiology 78 (2019) 143–154

Fig. 3. Effect of H2O2 on ROS accumulation as evaluated in six yeast strains with and without stress. The flow cytometry histogram profile expressed as the numberof events with 0 mM ( ) and 2 mM ( ) of H2O2. The mean fluorescence index (MFI) and ± SD of n = 3 was calculated according to Boettiger et al. (2001) as follows: [(geometric mean of the positive fluorescence) – (geometric mean of the control)]/(geometric mean of the control). Different letters in superscripts (atoe)indicate values of MFI significantly different between strains by Tukey's post-test (P <0.05). species could modify their lipid composition to better resist the oxi- compared to the other studied strains (approximately 15–25 μg sterol/ dative stress exposure. mg total protein). The primary sterol in all the strains was ergosterol, although the percentages varied markedly between species (38–96%). In the M. pulcherrima strains, practically the only sterol that was 3.3.1. Fatty acid, phospholipid and sterol composition before stress exposure quantified was ergosterol. Instead, the T. delbrueckii strains exhibited FAs typically make up parts of complex lipids, and they are im- the lowest percentage of ergosterol (38–40%) and the highest levels of portant structural components of biological membranes. FA analysis of squalene (33–36%) and lanosterol (6–11%). For S. cerevisiae, the strains the total cell extracts showed differences between the species (Table 1). used in this study showed significant differences in their sterol com- In the S. cerevisiae strains (QA23 and BY4742), MUFAs (palmitoleic positions. Thus, without accounting for the ergosterol, QA23 had a (C16:1) and oleic (C18:1) acids) and palmitic acid (C16:0) represented higher percentage of squalene and zymosterol, whereas BY4742 con- almost 90% of the FA in the cell extracts. By contrast, the non-Sac- tained a higher proportion of fecosterol than the other strains. charomyces strains contained a lower percentage of C16:1 (especially in the M. pulcherrima strains), which was compensated by the presence of linoleic acid (C18:2), a PUFA. Moreover, in the case of the M. pulcher- 3.3.2. Fatty acid, phospholipid and sterol composition after stress exposure rima strains, a low percentage of linolenic acid (C18:3) was also present. Differences in the cellular lipid compositions before and after stress As a result of this fatty acid pattern, the T. delbrueckii strains presented exposure are shown in Fig. 4. QA23 and BY4742 showed only a few higher UFA/SFA ratios than the other studied species. changes in the FA composition (Fig. 4). In QA23, slightly decreased PLs are major structural components of cell membranes and are amounts of C16:0 and C18:0 and increased amounts of C16:1 were essential for vital cellular processes. The PL percentages of the homo- found, leading to an increase in the unsaturated FA/saturated FA (UFA/ genates showed a similar composition in all the studied yeasts, with SFA) ratio and in the unsaturation index (UI) (Fig. 4). By contrast, non- phosphatidylcholine (PC) and phosphatidylethanolamine (PE) re- Saccharomyces species experienced highly modified FA compositions presenting approximately 50% and 24% of the total PLs, respectively after stress. The percentage of PUFAs (C18:2, and for Mp also C18:3) (Table 1). However, there were also small shifts in some PLs between and SFA (C16:0 and C18:0) decreased, whereas the MUFAs (C16:1 and different yeast species. In general, all the non-Saccharomyces strains C18:1) strongly increased. In fact, under these stress conditions, the showed a significantly lower percentage of dimethyl phosphatidy- percentages of C18:1 in non-Saccharomyces strains were higher than lethanolamine (DMPE), and the T. delbrueckii strains had the lowest they were in S. cerevisiae (Table 2), unlike what we observed under the amounts of lysophospholipids (LP). Of the strains studied here, Mpp control conditions. These variations resulted in a higher UFA/SFA ratio showed the most different PL composition, resulting in the highest PC/ and a lower UI in non-Saccharomyces species (Fig. 4 and Table 2). PE and the lowest phosphatidylinositol/phosphatidylserine (PI/PS) ra- The PL composition was slightly affected by stress, but the total PL tios. In fact, the highest PI/PS ratio was found in the QA23 strain. profile remained similar between species (Table 2). The PC and PE Sterols are essential lipid constituents present in yeasts as free persisted as the primary PLs in all the yeast strains, and although all the membranous sterols and as steryl esters. Although only the free sterol strains showed increased PC/PE ratios after stress (Fig. 4), this increase fraction is important for membrane properties or functions, esterifica- was statistically higher in the TdB and Mpp strains (Table 2). Moreover, tion of sterols helps to keep the level of free sterols balanced (Korber the PI/PS ratio decreased in non-Saccharomyces strains but increased et al., 2017). The total sterol content (Table 1) was significantly lower greatly in S. cerevisiae (Fig. 4) due to the increased PI and the decreased in both M. pulcherrima strains (4–8 μg sterol/mg total protein) PS in QA23 and BY4742. Notably, there was a significant decrease in

147 J. Vázquez et al. Food Microbiology 78 (2019) 143–154

Table 1 Fatty acids (FAs), phospholipids (PLs) and sterol composition of different strains after 6 h of growth in YPD medium. QA23 and BY4742 strains belongto S. cerevisiae species, TdB and Tdp strains to T. delbrueckii, and MpF and Mpp strains to M. pulcherrima. Different letter superscripts (a to e) indicate, for each studied compound, values significantly different between strains, by Tukey's post-test (P <0.05).

% Lípid composition Yeast strain

QA23 BY4742 TdB Tdp MpF Mpp

Fatty acids (FAs) Myristic (C14:0) acid 2.58 ± 0.34a 2.45 ± 0.59a 1.30 ± 0.34b 1.55 ± 0.09b 0.98 ± 0.02c 0.89 ± 0.30c Palmitic (C16:0) acid 23.17 ± 1.75a 23.55 ± 0.67a 18.30 ± 0.30b 20.03 ± 0.26c 23.27 ± 1.47a 21.76 ± 0.80a,d Palmitoleic (C16:1) acid 36.29 ± 3.14a 42.64 ± 0.63b 21.14 ± 2.28c 19.81 ± 0.21c 2.95 ± 0.09d 3.69 ± 0.72d Stearic (C18:0) acid 7.64 ± 1.57a 8.36 ± 1.16a 8.09 ± 0.19a 7.40 ± 0.56a 9.37 ± 0.34a 8.62 ± 0.19a Oleic (C18:1) acid 30.32 ± 0.53a 22.99 ± 1.14b 24.21 ± 0.52b 25.59 ± 0.86b,c 17.21 ± 0.61d 26.43 ± 1.10b,c Linoleic (C18:2) acid n.d. n.d. 26.95 ± 2.21a 25.63 ± 0.26a 40.17 ± 1.29b 34.67 ± 1.05d Linolenic (C18:3) acid n.d. n.d. n.d. n.d. 6.05 ± 0.16a 3.95 ± 0.15b Total FAs # 86.25 ± 1.14a 87.84 ± 1.48a 76.53 ± 9.34a 91.9 ± 2.72a,b 65.75 ± 6.17c 80.49 ± 0.80a,d

C16:1/C18:1 1.20 ± 0.08a 1.86 ± 0.12b 0.87 ± 0.11c 0.77 ± 0.03c 0.17 ± 0.00d 0.14 ± 0.03e UFA/SFA 2.01 ± 0.33a 1.91 ± 0.04a 2.61 ± 0.06b 2.45 ± 0.05b 1.98 ± 0.16a 2.20 ± 0.13a

∗ UI 0.67 ± 0.03a 0.66 ± 0.00a 0.99 ± 0.03b 0.97 ± 0.00b 1.19 ± 0.03c 1.11 ± 0.02d

Phospholipids (PLs) PI (Phosphatidylinositol) 12.49 ± 1.31a 12.19 ± 0.61a 14.63 ± 2.01a 12.93 ± 1.13a 10.10 ± 0.46b 8.94 ± 1.44b PS (Phosphatidylserine) 4.01 ± 0.21a 6.19 ± 0.76b 5.82 ± 1.61a,b 5.22 ± 1.29a,b 6.45 ± 1.42b 6.91 ± 0.63b,c PC (Phosphatidylcoline) 43.47 ± 0.51a 40.63 ± 1.07b 44.23 ± 1.55a 46.65 ± 2.00a 44.10 ± 4.56a 51.53 ± 0.93c PE (Phosphatidylethanolamine) 24.61 ± 1.38a 24.07 ± 1.66a 23.34 ± 2.64a 23.32 ± 4.19a 21.71 ± 0.17a,b 20.39 ± 1.03b CL (Cardiolipin) 5.66 ± 0.23a 2.86 ± 0.18b 6.02 ± 1.33a 6.87 ± 1.24a 7.03 ± 0.23a,c 4.71 ± 0.94a DMPE (Dimethyl- 3.76 ± 1.12a 5.39 ± 0.10b 0.99 ± 0.25c 1.18 ± 0.62c 1.60 ± 0.37c 1.02 ± 0.36c phosphatidylethanolamine) PA (Phosphatidic acid) 2.00 ± 0.28a 4.48 ± 2.57a 4.18 ± 1.27a,b 3.39 ± 1.37a 4.67 ± 0.67a,b 3.75 ± 0.11a,b LP (Lysophospholipids) 4.01 ± 0.21a 4.19 ± 2.06a 0.79 ± 0.25b 0.43 ± 0.54b 4.35 ± 1.48a 2.75 ± 0.29a,c

PI/PS 3.11 ± 0.16a 2.00 ± 0.34b 2.56 ± 0.36b 2.52 ± 0.40b 1.61 ± 0.43b,c 1.31 ± 0.33c PC/PE 1.77 ± 0.12a 1.69 ± 0.16a 1.90 ± 0.14a 2.02 ± 0.27a 2.03 ± 0.19a 2.53 ± 0.17b

Sterols Squalene 18.51 ± 2.41a 2.91 ± 0.52b 33.15 ± 3.19c 36.37 ± 3.93c 6.24 ± 0.09d n.d. Zymosterol 16.78 ± 0.80a 8.39 ± 0.29b 4.16 ± 0.56c 6.15 ± 1.19d 0.91 ± 0.33e n.d. 4-methylzymosterol 1.21 ± 0.08a n.d. n.d. 1.71 ± 0.87a n.d. n.d. Fecosterol 6.48 ± 0.05a 14.11 ± 0.32b 9.36 ± 0.37c 6.02 ± 0.07d n.d. 2.24 ± 1.16e 14-methylfecosterol n.d. n.d. n.d. 1.17 ± 0.10 n.d. n.d. Episterol 1.38 ± 0.06a n.d. 1.08 ± 0.07b 4.51 ± 0.89c n.d. n.d. Lanosterol 2.75 ± 0.04a 3.72 ± 0.22b 11.26 ± 1.31c 5.67 ± 0.89d 3.91 ± 1.32a,b,d 1.22 ± 1.03e Ergosterol 52.89 ± 1.58a 70.88 ± 0.77b 40.99 ± 2.00c 38.39 ± 1.70c 88.95 ± 1.56d 96.54 ± 1.89d Total sterols # 26.92 ± 0.69a 16.52 ± 0.40b 27.94 ± 0.87a 24.97 ± 4.69a 3.78 ± 0.86c 8.16 ± 0.50d

Ergosterol/Squalene 2.88 ± 0.46a 24.78 ± 4.71b 1.24 ± 0.18c 1.07 ± 0.16c 13.95 ± 0.05d –

# (μg/mg protein). * UI, unsaturation index. The unsaturation index was defined as follows: ((percentage of C16:1 + percentage of C18:1) + 2 (percentage of C18:2) + 3(percentage of C18:3))/100 (Rodríguez-Vargas et al., 2007). the cardiolipin (CL) content under stress exposure in all the strains component (Fig. 4). except for Mpp, the most H2O2-resistant strain, which increased the CL content under these conditions (Fig. 4). Moreover, both M. pulcherrima strains showed significantly decreased amounts of lysophospholipids 3.4. Principal component analysis (PCA) (LP), whereas the S. cerevisiae strains, especially BY4742, were the strains with higher LP content after stress (Table 2). PCA was applied to correlate the different variables (lipid compo- After the stress, ergosterol remained the primary sterol in all the sition, inhibition haloes and ROS levels (MFI)) and highlight some studied yeasts and the only one in Mpp (Table 2). However, a different grouping patterns within the different species under different condi- behavior was observed between the wine yeast strains and laboratory tions. Before stress (Fig. 5A), the species were clearly separated into strain BY4742. All the wine yeasts showed increased ergosterol con- three groups by their lipid composition (Table 1), with M. pulcherrima tents and decreased squalene contents under stress (resulting in an in- being the most diverse compared to S. cerevisiae and T. delbrueckii. Both crease of the ergosterol/squalene ratio; Fig. 4). However, BY4742 M. pulcherrima strains (MpF and Mpp) were different from the other showed the opposite behavior, with increasing squalene and decreasing strains in that they exhibited higher ergosterol and PS percentages and ergosterol contents, resulting in a decrease in the ergosterol/squalene lower PI and PE. Furthermore, the percentage of total PUFAs (C18:2 ratio (Fig. 4). In fact, BY4742 showed the highest value for this ratio and C18:3) was clearly higher in the M. pulcherrima strains (the T. before stress and the lowest after stress (Table 2). However, the T. delbrueckii only showed low levels of C18:2, and S. cerevisiae had no delbrueckii strains showed the highest ergosterol/squalene ratios under PUFAs in its lipid composition). Both S. cerevisiae strains were char- stress (mostly due to the drop in squalene content), and they were the acterized by high levels of myristic acid (C14:0) and oleic acid (C16:1), strains that had more diverse sterol compounds and the only species DMPE, PI/PS ratios and zymosterol and low UI values. The lowest LP, that exhibited methyl fecosterol. In fact, whereas both S. cerevisiae C16:0 and ergosterol contents and the highest squalene content and strains experienced decreases in their zymosterol and fecosterol per- UFA/SFA ratio were characteristics of the T. delbrueckii species, which centages under stress, the T. delbrueckii strains increased their showed similarities with the other non-Saccharomyces but also with Saccharomyces.

148 J. Vázquez et al. Food Microbiology 78 (2019) 143–154

Fig. 4. Heatmap representing the fold changes in the lipid composition of cells following stress exposure to the cells before stress exposure. S. cerevisiae strains: QA23 and BY4742; T. delbrueckii strains: TdB and Tdp; and M. pulcherrima strains: MpF and Mpp.

The key indicative features of oxidative stress were the higher ROS primary differences were already observed before stress exposure (Fig. accumulation (Fig. 3) and higher inhibition haloes (Fig. 2), which were S1 and Table S2). positively correlated with the percentage of C14:0, stearic acid (C18:0), C16:1, PI, PE, DMPE, LP and squalene, and the PI/PS and C16:1/C18:1 ratios (Fig. 5B, positive component 1 and Table S1). However, high ROS 4. Discussion and inhibition haloes were negatively correlated with the UFA (C18:1, C18:2 and C18:3), UI, CL, PC and ergosterol contents, and with the Although S. cerevisiae is the wine yeast par excellence due to its ratios of PC/PE and ergosterol/squalene (Fig. 5B, negative component 1 fermentative capacity, there is currently a strong interest that is being and Table S1). Thus, both S. cerevisiae strains were clearly different driven by consumer and industry demand for wines with improved from both non-Saccharomyces species because they were grouped on the characteristics to study the possibility of using non-conventional yeasts positive side of component 1, which is indicative of less stress tolerance with peculiar features in industrial fermentations. Under these condi- (with BY4742 having higher positive values). M. pulcherrima strains tions, yeasts are exposed to a variety of stresses, such as the oxidative were placed on the opposite side (negative component 1), with Mpp stress. For instance, S. cerevisiae, used as starter in biotech and food being the strain that exhibited more negatives values, indicating a industries in active dry yeast (ADY) form, can suffer oxidative stress higher resistance to stress. Thus, the component 1 places the strains during the biomass propagation and dehydration steps of their pro- according to their resistance to oxidative stress, with the less H2O2- duction, which could negatively affect yeast performance (reviewed by tolerant strain BY4742 on one side, and Mpp on the opposite/negative Matallana and Aranda, 2017). The oxidative stress response is a po- side, which was the most resistant to this stress. tential target for wine yeast improvement (Gamero-Sandemetrio et al., Although the differences between species in terms of lipid compo- 2014). Many studies on stress resistance have been performed in S. sition increased under stress, it is important to highlight that the cerevisiae, but few have addressed other yeast species, which have also shown a significant impact on food and beverage production (Pretorius,

149 J. Vázquez et al. Food Microbiology 78 (2019) 143–154

Table 2 Fatty acids (FAs), phospholipids (PLs) and sterols composition of different strains after stress exposure at 24 h of growth in YPD medium. QA23 and BY4742strains belong to S. cerevisiae species, and TdB, and Tdp strains to T. delbrueckii and MpF and Mpp strains to M. pulcherrima. Different letter superscripts (a to e) indicate, for each studied compound, values significantly different between strains (by Tukey's post-test, P < 0.05), and asterisk indicate values significantly differentbetween cells before (Table 1) and after stress exposure (by Tukey's post-test, P < 0.05).

% Lípid composition Yeast strain

QA23 BY4742 TdB Tdp MpF Mpp

Fatty acids (FAs) Myristic (C14:0) acid 2.03 ± 0.71a 2.35 ± 0.35a 1.83 ± 0.19a 1.68 ± 0.09a,b 1.12 ± 0.07c 0.68 ± 0.16d ∗ ∗ ∗ ∗ Palmitic (C16:0) acid 19.88 ± 0.20a, 23.63 ± 0.73b 13.94 ± 2.11c, 15.18 ± 0.08c, 21.75 ± 0.93b 17.17 ± 0.63c, ∗ ∗ ∗ ∗ ∗ Palmitoleic (C16:1) acid 40.44 ± 0.33a, 43.28 ± 0.89b 33.32 ± 0.58c, 31.07 ± 0.02d, 10.81 ± 0.45e, 8.59 ± 0.29f, ∗ ∗ ∗ ∗ Stearic (C18:0) acid 6.61 ± 0.86a 8.47 ± 1.05a 3.91 ± 0.41b, 4.03 ± 0.02b, 4.14 ± 0.61b, 3.48 ± 0.08b,c ∗ ∗ ∗ ∗ Oleic (C18:1) acid 30.53 ± 0.32a 22.27 ± 1.25b 39.96 ± 1.24c, 41.98 ± 0.73c, 31.28 ± 1.16a, 52.23 ± 0.56d, ∗ ∗ ∗ ∗ Linoleic (C18:2) acid n.d. n.d. 7.03 ± 0.89a, 6.07 ± 0.76a, 30.00 ± 0.80b, 17.32 ± 0.39c, ∗ ∗ Linolenic (C18:3) acid n.d. n.d. n.d. n.d. 0.90 ± 0.16a, 0.54 ± 0.05b, ∗ ∗ ∗ ∗ Total FAs # 86.26 ± 3.70a 68.12 ± 4.15b, 101.08 ± 0.43c, 93.42 ± 3.53a 88.02 ± 0.34a, 74.2 ± 1.43d,

∗ ∗ C16:1/C18:1 1.33 ± 0.00a, 1.95 ± 0.07b 0.83 ± 0.01d 0.74 ± 0.01e 0.35 ± 0.03f, 0.16 ± 0.00g ∗ ∗ ∗ ∗ ∗ UFA/SFA 2.49 ± 0.02a, 1.91 ± 0.18b 4.13 ± 0.71c, 3.79 ± 0.00c, 2.70 ± 0.03d, 3.69 ± 0.16c,

∗ ∗ ∗ ∗ ∗ ∗ UI 0.71 ± 0.01a, 0.66 ± 0.02a 0.87 ± 0.04b, 0.85 ± 0.01b, 1.05 ± 0.01c, 0.97 ± 0.01d,

∗ ∗ Phospholipids PI (Phosphatidylinositol) 15.86 ± 3.40a 15.83 ± 2.37a, 14.52 ± 2.82a 13.48 ± 0.50a 7.71 ± 0.46b, 8.37 ± 0.19a ∗ (PLs) PS (Phosphatidylserine) 3.46 ± 0.21a, 5.09 ± 0.63b 7.32 ± 1.69b 5.76 ± 1.65a,b 6.35 ± 0.54b 7.37 ± 0.42b ∗ ∗ ∗ PC (Phosphatidylcoline) 46.55 ± 2.85a, 41.38 ± 1.80b 45.90 ± 2.37a 51.73 ± 1.87b, 51.12 ± 1.69b, 53.90 ± 3.05b ∗ ∗ PE (Phosphatidylethanolamine) 23.06 ± 0.54a 23.82 ± 0.67a 19.73 ± 2.88b, 21.36 ± 2.97a,b 22.13 ± 1.11b 17.01 ± 1.01c, ∗ ∗ ∗ ∗ ∗ CL (Cardiolipin) 3.44 ± 0.45a, 1.57 ± 0.07b, 4.67 ± 0.12c, 4.29 ± 0.22c, 5.93 ± 0.50c, 6.64 ± 1.48c ∗ ∗ DMPE (Dimethyl- 2.34 ± 0.15a, 3.63 ± 0.82b, 1.61 ± 0.73c 1.28 ± 0.28c 1.10 ± 0.44c 1.62 ± 0.74c phosphatidylethanolamine) PA (Phosphatidic acid) 2.87 ± 0.60a 4.54 ± 0.76a,b 3.97 ± 2.08a,b 1.33 ± 1.37a 3.90 ± 0.70a,b 3.65 ± 0.19a,b ∗ ∗ ∗ LP (Lysophospholipids) 2.42 ± 0.27a, 4.14 ± 0.11b 1.28 ± 1.07c 0.75 ± 0.80c 1.77 ± 0.06c, 1.44 ± 0.16c,

∗ ∗ ∗ PI/PS 4.64 ± 0.35a, 3.10 ± 0.26b, 1.99 ± 0.07c, 2.33 ± 0.13d 1.21 ± 0.03e 1.13 ± 0.09e ∗ ∗ ∗ PC/PE 2.02 ± 0.08a, 1.74 ± 0.03b 2.36 ± 0.16c, 2.40 ± 0.22c 2.32 ± 0.19c 3.18 ± 0.36d,

∗ ∗ ∗ ∗ ∗ Sterols Squalene 6.37 ± 1.75a, 11.01 ± 0.86b, 3.06 ± 0.15c, 3.34 ± 1.52c, 2.51 ± 0.49c, n.d. ∗ ∗ ∗ ∗ ∗ Zymosterol 13.90 ± 0.57a, 6.10 ± 1.16b, 11.48 ± 1.56a, 9.09 ± 0.27c, 1.28 ± 0.02d, n.d. ∗ ∗ ∗ 4-methylzymosterol 0.90 ± 0.07a, n.d. 2.29 ± 0.53b, 2.77 ± 0.86b, n.d. n.d. ∗ ∗ ∗ ∗ Fecosterol 4.06 ± 0.17a, 11.09 ± 1.30b, 11.78 ± 0.21b, 9.63 ± 0.77b, n.d. n.d. ∗ ∗ 14-methylfecosterol n.d. n.d. 1.91 ± 0.34a, 1.77 ± 0.38a, n.d. n.d. ∗ ∗ ∗ Episterol 0.31 ± 0.44a, n.d. 1.59 ± 0.08b, 1.73 ± 0.30c, n.d. n.d. ∗ ∗ ∗ ∗ Lanosterol 1.63 ± 0.17a, 6.05 ± 0.34b, 7.88 ± 0.23c. 8.97 ± 1.38c, 1.76 ± 0.59a n.d. ∗ ∗ ∗ ∗ ∗ ∗ Ergosterol 72.83 ± 1.48a, 65.15 ± 1.94b, 60.89 ± 3.34b, 62.69 ± 2.81b, 94.47 ± 1.69c, 100 ± 0.00d, ∗ ∗ ∗ ∗ ∗ Total sterols # 40.91 ± 2.10a, 20.00 ± 2.17b, 43.41 ± 2.48a, 27.80 ± 1.60c 8.52 ± 1.00d, 9.99 ± 1.02d,

∗ ∗ ∗ ∗ ∗ Ergosterol/Squalene 11.90 ± 3.49a, 5.62 ± 0.71b, 19.92 ± 2.04c, 21.13 ± 2.46c, 38.26 ± 2.46d, –

# (μg/mg protein). * UI, unsaturation index. The unsaturation index was defined as follows: ((percentage of C16:1 + percentage of C18:1) + 2 (percentage of C18:2) + 3(percentage of C18:3))/100 (Rodríguez-Vargas et al., 2007).

2000). In this study, we evaluated oxidative stress tolerance in selected 1998) with very poor fermentation capacity (Rossouw et al., 2013). non-Saccharomyces wine strains, namely, T. delbrueckii and M. pulcher- Unlike wine yeast strains, BY4742 is not adapted to withstand adverse rima, and we compared it to the S. cerevisiae response. growth conditions such as those found during the fermentation process Our findings clearly indicated that these non-conventional yeasts (Carrasco et al., 2001). are more tolerant to external oxidative stress than S. cerevisiae. As re- The cell's first barrier against stress is the cellular membrane, and ported elsewhere (Jamieson, 1998; Moradas-Ferreira et al., 1996; lipids are one of its primary components. In this study, we evaluated the Moradas-Ferreira and Costa, 2000), exposing yeast to H2O2 was asso- differences in lipid composition between the species before andafter ciated with a rapid ROS generation and a loss of viability, at least until stress exposure. Our results showed that the cellular lipid composition the yeast manages to adapt to the new environmental conditions, i.e., differed widely between species, and thus it may be involved withtheir after the activation of defense mechanisms to maintain a proper redox different abilities to resist and tolerate oxidative stress. Regardless of state. Under our conditions, the M. pulcherrima species, and especially stress, the primary feature was the high fatty acid unsaturation rate the autochthonous strain (Mpp), exhibited the greatest resistance to observed in both non-Saccharomyces species, which was basically due to oxidative stress (low ROS generation and higher H2O2 tolerance). Both the presence of PUFAs, resulting in high membrane fluidity. It is well T. delbrueckii strains, also showed a higher oxidative resistance com- known that S. cerevisiae cannot synthesize PUFAs because it only con- pared with S. cerevisiae as reported by Alves-Araújo et al. (2004) in a tains one desaturase, Δ9 fatty acid desaturase (OLE1), which can only baking industry study. Furthermore, all the wine yeasts tested here produce MUFAs of 16- and 18-carbon compounds (Stukey et al., 1990). were clearly more resistant to oxidative stress than the laboratory However, S. cerevisiae can incorporate exogenous PUFAs into its cell strain, probably due to their adaptive evolution to adverse stress con- membranes (Beltran et al., 2008; Rosi and Bertuccioli, 1992). Instead, ditions (Guillamón and Barrio, 2017; Querol et al., 2003). In fact, the in yeasts such as Kluyveromyces lactis, oleic acid (C18:1) is subsequently BY4742 strain grew poorly after stress was applied, achieving less than desaturated to linoleic acid (C18:2) and then to α-linolenic acid (C18:3) one more generation after stress exposure in liquid medium. BY4742 is by Δ12 and omega (Δ15) fatty acid desaturases, respectively (Ratledge part of a set of deletion strains derived from S288C (Brachmann et al., and Evans, 1989; Kainou et al., 2006; Santomartino et al., 2017). In our

150 J. Vázquez et al. Food Microbiology 78 (2019) 143–154

Fig. 5. Biplots of principal components analysis (PCA) using fatty acids (FAs), phospholipids (PLs), sterols, inhibition halo measures and ROS accumu- lation markers (MFI) as variables. S. cerevisiae strains: QA23 ( ) and BY4742 ( ); T. delbrueckii strains: TdB ( ) and Tdp ( ); M. pulcherrima strains: MpF ( ) and Mpp ( ). The explicative variables were distributed along the PCA as follows: (A) Biplot with 72.60% of the variance before the oxidative stress was applied. Component 1: (+); phosphatidylinositol (PI), phosphatidylethanolamine, (PE), PI/phosphati- dylserine (PI/PS) ratio, mirystic (C14:0) and palmi- toleic (C16:1) acids, C16:1/oleic (C16:1/C18:1) ratio, fecosterol and zymosterol. (−); PS, phosphatidylco- line (PC), PC/PE ratio, stearic (C18:0), linoleic (C18:2) and linolenic (C18:3) acids, ergosterol, un- saturated index and ergosterol/squalene ratio. Component 2: (+); unsaturated/saturated (UFA/ SFA) ratio, cardiolipin (CL), squalene, 14-methylfe- costerol and episterol. (−); dimethylpho- sphatidylethanolamine (DMPE), lysophospholipids (LP) and palmitic (C16:0) acid. (B) Biplot with 82.18% of the variance after oxidative stress (2 mM

H2O2) was applied. Component 1: (+); inhibition halos, MFI, C14:0, C16:0, C16:1, C18:0, C16:1/C18:1 ratios, squalene, fecosterol, zymosterol, PI, PE, DMPE and PI/PS. (−); C18:1, C18:2, C18:3, ergosterol, er- gosterol/squalene ratio, PC, CL, PC/PE ratio and unsaturation index. Component 2: (+); 4-methilzy- mosterol, 14-methylfecosterol, episterol, lanosterol and UFA/SFA ratio. (−); LP and C16:0.

case, both non-Saccharomyces species presented PUFAs in their lipid although PUFAs seem to increase yeast tolerance to stress, they can also compositions, although linolenic acid (C18:3) was a unique feature of be toxic to cells because of their susceptibility to peroxidation (Cipak the M. pulcherrima strains. By contrast, the ratio C16:1/C18:1 ratio was et al., 2006; Johansson et al., 2016). In fact, the heterologous produc- higher in the S. cerevisiae strains, with the highest content of palmitoleic tion of PUFAs in S. cerevisiae has been shown to increase oxidative stress acid (C16:1), the primary UFA in aerobically grown S. cerevisiae strains (Ruenwai et al., 2011), and in non-Saccharomyces strains, a higher (Steels et al., 1994), which is correlated with higher membrane rigidity proportion of C18:2 acid does not assure increased tolerance to ethanol (Redón et al., 2009). Many S. cerevisiae studies have reported a corre- stress (Aguilera et al., 2006; Archana et al., 2015). Under our condi- lation between an increase in membrane fluidity (due to an increase in tions, i.e., under oxidative stress, high levels of C18:2 acid were posi- the degree of unsaturation) and a higher tolerance to various types of tively correlated with low ROS generation and high H2O2 tolerance. stresses, such as cold or ethanol stress (Beltran et al., 2008; Casey and Nevertheless, the amounts of PUFAs decreased in all the non-Sacchar- Ingledew, 1986; Guerzoni et al., 1997; Suutari and Laakso, 1994). omyces strains after stress exposure, probably indicating that the Therefore, according to the unsaturation degree, the studied non-Sac- strategy of these species was a reduction of the PUFA content due to charomyces species were positively correlated with higher oxidative their high sensitivity to peroxidation (Ayala et al., 2014; Johansson stress resistance. In fact, the introduction of the gene encoding the Δ12 et al., 2016). This effect could be a mechanism in non-conventional fatty acid desaturase gene (FAD2) in the S. cerevisiae strains reportedly yeasts to withstanding the oxidative stress better without compromising resulted in a higher resistance to ethanol (Kajiwara et al., 1996), and membrane integrity. The other principal mechanism used by non-Sac- NaCl and freezing (Rodríguez-Vargas et al., 2007). Moreover, the in- charomyces yeasts to cope with oxidative stress was the modulation of troduction of both desaturases (FAD2 and FAD3 (ω3 fatty acid desa- their FA composition, by raising the proportion of MUFAs, such as turase) from K. lactis) into a strain of S. cerevisiae has been reported to palmitoleic acid and oleic acid, and by decreasing the amounts of SFA, increase the alkaline pH tolerance (Yazawa et al., 2009). Therefore, such as palmitic acid and stearic acid. Oleic acid has been suggested as

151 J. Vázquez et al. Food Microbiology 78 (2019) 143–154 a membrane fluidity sensor, and it seems to be the most important UFA 2014). Although M. pulcherrima showed the highest ergosterol percen- for counteracting the toxic nature of ethanol by increasing the mem- tage (but the lowest content), this parameter could not be correlated brane stability and antagonizing the fluidity caused by ethanol (You with the oxidative stress either. Nevertheless, M. pulcherrima might be et al., 2003). Furthermore, palmitoleic acid is induced by stress in high- compensating for fluidization effect elicited by the oxidative stress. The density fermentations, and it has a protective function against damage overall fluidity/rigidity of a membrane is the result of a combination of (Ding et al., 2009). According to Redón et al. (2009), the supple- all parameters mentioned (Ding et al., 2009). Unsaturated fatty acids, mentation of palmitoleic acid in wine yeast culture has a positive effect such as oleic acid, increase cell membrane fluidity. On the other hand, on the yeast viability and the fermentation kinetics. However, although low amounts of ergosterol would facilitate membrane rigidity, allowing the UFA/SFA ratio increased, the results showed a decrease in the un- yeast to maintain membrane functionality. saturation index in non-Saccharomyces species, indicating how yeasts The growth of the BY4742 strain was clearly affected after stress try to maintain their membrane fluidity. exposure, and it showed the highest squalene content and the lowest Regarding the phospholipid composition, PC and PE are the primary ergosterol/squalene ratio after stress exposure, being both parameters PLs of yeast membranes, representing up to 60–70% of total PLs positively correlated with less tolerance to stress. It has to be taken into (Schneiter et al., 1999). The PC/PE ratio is an important parameter for account that squalene, the precursor of the synthesis of ergosterol, is a the biophysical status of the membrane (Flis et al., 2015). An increase highly hydrophobic molecule, which lacks the amphipathic character in the PC/PE ratio has been reported as one of the yeast adaptation provided by the hydroxyl group at C3 atom present in sterols. However, mechanisms to oxidative stress, leading to a reorganization of the even if under standard conditions it is stored in lipid droplets, it can plasma membrane lipid composition, and a decrease of the membrane also be found in organelle membranes (e.g in yeast cells grown anae- permeability against H2O2 (Pedroso et al., 2009). On the other hand, a robically or in strains lacking HEM1 gene) without causing deleterious decrease of PC/PE ratio, in combination with low UFA/SFA and high effects (Spanova et al., 2010, 2012). Therefore, although squalene is not amounts of ergosterol, leads to higher transition temperature of a lipid a typical membrane lipid, it may be considered as a mild modulator of bilayer, and membranes may become more rigid (Flis et al., 2015). biophysical membrane properties (Spanova et al., 2012). Thus, the large quantities of PC in non-conventional yeasts could lead to High ratios of UFA/SFA and high PC/PE ratios in membranes are a decrease of cellular permeability to H2O2 (Pedroso et al., 2009), en- known to lead to high membrane fluidity (Flis et al., 2015). In the hancing oxidative stress tolerance, in a similar way to what has been natural strains of this study, these parameters also seem to lead to observed in S. cerevisiae ethanol tolerance (Chi and Arneborg, 1999; higher tolerance against to H2O2. Vendramin-Pintar et al., 1995). The PI/PS ratio is another important parameter for cell growth potential and essential for maintaining cel- 5. Conclusions lular viability in S. cerevisiae (Xia et al., 2011). The synthesis of these PLs is closely correlated, because both require the same precursor cy- In conclusion, our results suggest that non-conventional yeasts are tidyldiphosphate diacylglycerol (CDP-DAG) precursor. Moreover, PS best at resisting induced oxidative stress. The highest stress tolerance can be a precursor for the synthesis of PE and PC (Voelker and Frazier, was associated with the non-conventional yeasts' abilities to maintain a 1986). However, PI is considered essential for S. cerevisiae because the high proportion and level of unsaturated fatty acids, particularly lino- lack of this PL can reduce cell viability (Becker and Lester, 1977; De lenic acid and linoleic acid. Furthermore, the large variability in the Kroon et al., 2013). Our results show that the S. cerevisiae strains ex- fatty acid composition can result from adaptive responses to changes in hibited a high PI/PS ratio, especially after stress exposure, whereas the external physico-chemical parameters. M. pulcherrima strains, which had the highest resistance to H2O2 stress, exhibited the lowest values, especially after stress exposure. In fact, this Acknowledgements ratio was negatively correlated with tolerance to oxidative stress. Mitochondria are both the source and the site for the detoxification The authors thank the Ministry of Economy and Competitiveness, of reactive oxygen species in yeast (Chevtzoff et al., 2010; Rhoads et al., Spain (Projects AGL2013-47300-C3-1-R and AGL2016-77505-C3-3-R), 2006). Therefore, normal mitochondrial function is required for re- for its financial support. Jennifer Vázquez is grateful for the pre-doc- sistance to oxidative stress (Grant et al., 1997), and the maintenance of toral fellowship from the University Rovira i Virgili and the Oenological a stable respiratory chain strongly prevents the generation of mi- Biotechnology research group for a mobility grant, and she thanks tochondrial ROS (Barros et al., 2003). CL, a mitochondrial PL found in Francesca DiBartolomeo for generously teaching laboratory skills and the inner mitochondrial membrane (De Kroon et al., 2013), plays a key some procedures in Graz. role in the stabilization of electron transport chain complexes and the resistance against oxidative stress during respiratory growth (Chen Appendix A. Supplementary data et al., 2008). Most of the strains in this study, and especially those of S. cerevisiae, decreased their CL content after stress exposure, which could Supplementary data to this article can be found online at https:// indicate a destabilization of the respiratory chain. However, despite this doi.org/10.1016/j.fm.2018.10.001. decrease, non-Saccharomyces species displayed higher CL values, sug- gesting a better maintenance of functional mitochondria during H2O2 References stress as previously described for ethanol stress (Chi and Arneborg, 1999). Acar, F., 1980. The disc susceptibility test. In: Baltimore, L.V. (Ed.), Antibiotics in Sterols are necessary for maintaining membrane integrity and es- Laboratory Medicine, London, pp. 24–54. Alves-Araújo, C., Almeida, M.J., Sousa, M.J., Leão, C., 2004. Freeze tolerance of the yeast sential for cell viability (Daum et al., 1998). Consistent with the results Torulaspora delbrueckii: cellular and biochemical basis. FEMS Microbiol. Lett. 240, obtained by Murakami et al. (1996) in freezing-tolerant strains, our 7–14. https://doi.org/10.1016/j.femsle.2004.09.008. Aguilera, F., Peinado, R.A., Millán, C., Ortega, J.M., Mauricio, J.C., 2006. Relationship study demonstrated that the sterol content was low in the most H2O2- between ethanol tolerance, H+ -ATPase activity and the lipid composition of the tolerant strains (M. pulcherrima species). However, a direct correlation plasma membrane in different wine yeast strains. Int. J. Food Microbiol. 110, 34–42. between the sterol content and H2O2 tolerance was not observed be- https://doi.org/10.1016/j.ijfoodmicro.2006.02.002. cause, e.g., BY4742 was the most sensitive strain, but it also contained Archana, K.M., Ravi, R., Anu-Appaiah, K.A., 2015. Correlation between ethanol stress and low levels of sterols. Ergosterol, the primary yeast sterol and the end cellular fatty acid composition of alcohol producing non-Saccharomyces in compar- ison with Saccharomyces cerevisiae by multivariate techniques. J. Food Sci. Technol. product of the yeast sterol biosynthetic pathway, is responsible for 52, 6770–6776. https://doi.org/10.1007/s13197-015-1762-y. structural membrane features such as higher fluidity and other phy- Athenstaedt, K., Zweytick, D., Jandrositz, A., Kohlwein, S.D., Daum, G., 1999. siological functions (Parks, 1978; Daum et al., 1998; Klug and Daum, Identification and characterization of major lipid particle proteins of theyeast

152 J. Vázquez et al. Food Microbiology 78 (2019) 143–154

Saccharomyces cerevisiae. J. Bacteriol. 181, 6441–6448. of total lipides from animal tissues. J. Biol. Chem. 226, 497–509. Avery, S.V., 2011. Molecular targets of oxidative stress. Biochem. J. 434, 201–210. Gamero-Sandemetrio, E., Gomez-Pastor, R., Matallana, E., 2014. Antioxidant defense https://doi.org/10.1042/BJ20101695. parameters as predictive biomarkers for fermentative capacity of active dried wine Ayala, A., Muñoz, M.F., Arguelles, S., 2014. Lipid peroxidation: production, metabolism, yeast. Biotechnol. J. 9, 1055–1064. https://doi.org/10.1002/biot.201300448. and signaling mechanisms of malondialdehyde and 4-hydroxy-2-nonenal. Oxid. Med. Gibson, B.R., Lawrence, S.J., Boulton, C.A., Box, W.G., Graham, N.S., Linforth, R.S.T., Cell. Longev. 2014. https://doi.org/10.1155/2014/360438. Smart, K.A., 2008. The oxidative stress response of a lager brewing yeast strain Bauer, A.W., Kirby, W.M.M., J.C, Turck, M., 1966. Antibiotic susceptibility testing by a during industrial propagation and fermentation. FEMS Yeast Res. 8, 574–585. standardized single disk method. Am. J. Clin. Pathol. 45, 493–496. https://doi.org/10.1111/j.1567-1364.2008.00371.x. Barros, M.H., Netto, L.E.S., Kowaltowski, A.J., 2003. H2O2 generation in Saccharomyces Gibson, B.R., Lawrence, S.J., Leclaire, J.P.R., Powell, C.D., Smart, K.A., 2007. Yeast re- cerevisiae respiratory pet mutants: effect of cytochrome c. Free Radic. Biol. Med. 35, sponses to stresses associated with industrial brewery handling. FEMS Microbiol. Rev. 179–188. 31, 535–569. https://doi.org/10.1111/j.1574-6976.2007.00076.x. Becker, G.W., Lester, R.L., 1977. Changes in phospholipids of Saccharomyces cerevisiae González-Royo, E., Pascual, O., Kontoudakis, N., Esteruelas, M., Esteve-Zarzoso, B., Mas, associated with inositol-less death. J. Biol. Chem. 252, 8684–8691. A., Canals, J.M., Zamora, F., 2015. Oenological consequences of sequential inocula- Belda, I., Ruiz, J., Esteban-Fernández, A., Navascués, E., Marquina, D., Santos, A., tion with non-Saccharomyces yeasts (Torulaspora delbrueckii or Metschnikowia pul- Moreno-Arribas, M.V., 2017. Microbial contribution to wine aroma and its intented cherrima) and Saccharomyces cerevisiae in base wine for sparkling wine production. use for wine quality improvement. Molecules 22, 1–29. https://doi.org/10.3390/ Eur. Food Res. Technol. 240, 999–1012. https://doi.org/10.1007/s00217-014- molecules22020189. 2404-8. Beltran, G., Novo, M., Guillamón, J.M., Mas, A., Rozès, N., 2008. Effect of fermentation Grant, C.M., Maclever, F.H., Dawes, I.W., 1997. Mitochondrial function is required for temperature and culture media on the yeast lipid composition and wine volatile resistance to oxidative stress in the yeast Saccharomyces cerevisiae. FEBS Lett. 410, compounds. Int. J. Food Microbiol. 121, 169–177. https://doi.org/10.1016/j. 219–222. https://doi.org/10.1016/S0014-5793(97)00592-9. ijfoodmicro.2007.11.030. Guerzoni, M.E., Ferruzzi, M., Sinigaglia, M., Criscuoli, G.C., 1997. Increased cellular fatty Beltran, G., Torija, M.J., Novo, M., Ferrer, N., Poblet, M., Guillamon, J.M., Rozes, N., Mas, acid desaturation as a possible key factor in thermotolerance in Saccharomyces cere- A., 2002. Analysis of yeast populations during alcoholic fermentation: a six year visiae. Can. J. Microbiol. 43, 569–576. follow-up study. Syst. Appl. Microbiol. 25, 287–293. https://doi.org/10.1078/0723- Guillamón, J.M., Barrio, E., 2017. Genetic polymorphism in wine yeasts: mechanisms and 2020-00097. methods for its detection. Front. Microbiol. 8, 806. https://doi.org/10.3389/fmicb. Bely, M., Stoeckle, P., Masneuf-Pomarède, I., Dubourdieu, D., 2008. Impact of mixed 2017.00806. Torulaspora delbrueckii-Saccharomyces cerevisiae culture on high-sugar fermentation. Herrero, E., Ros, J., Bellí, G., Cabiscol, E., 2008. Redox control and oxidative stress in Int. J. Food Microbiol. 122, 312–320. https://doi.org/10.1016/j.ijfoodmicro.2007. yeast cells. Biochim. Biophys. Acta 1780, 1217–1235. https://doi.org/10.1016/j. 12.023. bbagen.2007.12.004. Beney, L., Gervais, P., 2001. Influence of the fluidity of the membrane on the response of Howlett, N.G., Avery, S.V., 1997. Induction of lipid peroxidation during heavy metal microorganisms to environmental stresses. Appl. Microbiol. Biotechnol. 57, 34–42. stress in Saccharomyces cerevisiae and influence of plasma membrane. Fatty Acid https://doi.org/10.1007/s002530100754. Unsaturation 63, 2971–2976. Boettiger, D., Huber, F., Lynch, L., Blystone, S., 2001. Activation of αvβ3-vitronectin Hunter, Rose, 1972. Lipid composition of Saccharomyces cerevisiae as influenced by binding is a multistage process in which increases in bond strength are dependent on growth temperature. Biochim. Biophys. Acta 260, 639–653. Y747 and Y759 in the cytoplasmic domain of β3. Mol. Biol. Cell 12, 1227–1237. Jamieson, D.J., 1998. Oxidative stress responses of the yeast Saccharomyces cerevisiae. https://doi.org/10.1091/mbc.12.5.1227. Yeast 14, 1511–1527. https://doi.org/10.1002/(SICI)1097-0061(199812) Brachmann, C.B., Davies, A., Cost, G.J., Caputo, E., Li, J., Hieter, P., Boeke, J.D., 1998. 14:16<1511::AID-YEA356>3.0.CO;2-S. Designer deletion strains. Yeast 14, 115–132. https://doi.org/10.1002/(SICI)1097- Johansson, M., Chen, X., Milanova, S., Santos, C., Petranovic, D., 2016. PUFA-induced cell 0061(19980130)14:2<115::AID-YEA204>3.0.CO;2-2. death is mediated by Yca1p-dependent and -independent pathways, and is reduced Broekhuyse, R.M., 1968. Phospholipids in tissues of the eye I. Isolation, characterization by vitamin C in yeast. FEMS Yeast Res. 16, 1–9. https://doi.org/10.1093/femsyr/ and quantitative analysis by two-dimensional thin-layer chromatography of diacyl fow007. and vinyl-ether phospholipids. Biochim. Biophys. Acta Lipids Lipid Metab. 152, Jolly, N.P., Varela, C., Pretorius, I.S., 2014. Not your ordinary yeast: non-Saccharomyces 307–315. https://doi.org/10.1016/0005-2760(68)90038-6. yeasts in wine production uncovered. FEMS Yeast Res. 14, 215–237. https://doi.org/ Carrasco, P., Querol, A., Del Olmo, M., 2001. Analysis of the stress resistance of com- 10.1111/1567-1364.12111. mercial wine yeast strains. Arch. Microbiol. 175, 450–457. https://doi.org/10.1007/ Kainou, K., Kamisaka, Y., Kimura, K., Uemura, H., 2006. Isolation of Δ12 and ω3-fatty s002030100289. acid desaturase genes from the yeast Kluyveromyces lactis and their heterologous Casey, G.P., Ingledew, W.M., 1986. Ethanol tolerance in yeasts. Crit. Rev. Microbiol. 13, expression to produce linoleic and α-linolenic acids in Saccharomyces cerevisiae. Yeast 219–280. https://doi.org/10.3109/10408418609108739. 23, 605–612. https://doi.org/10.1002/yea.1378. Chen, S., He, Q., Greenberg, M.L., 2008. Loss of tafazzin in yeast leads to increased Kajiwara, S., Shirai, A., Fujii, T., Toguri, T., Nakamura, K., Ohtaguchi, K., 1996. oxidative stress during respiratory growth. Mol. Microbiol. 68, 1061–1072. https:// Polyunsaturated fatty acid biosynthesis in Saccharomyces cerevisiae: expression of doi.org/10.1111/j.1365-2958.2008.06216.x. ethanol tolerance and the FAD2 gene from Arabidopsis thaliana. Appl. Environ. Chevtzoff, C., Yoboue, E.D., Galinier, A., Casteilla, L., Daignan-Fornier, B., Rigoulet, M., Microbiol. 62, 4309–4313. Devin, A., 2010. Reactive oxygen species-mediated regulation of mitochondrial bio- Klug, L., Daum, G., 2014. Yeast lipid metabolism at a glance. FEMS Yeast Res. 14, genesis in the yeast Saccharomyces cerevisiae. J. Biol. Chem. 285, 1733–1742. https:// 369–388. https://doi.org/10.1111/1567-1364.12141. doi.org/10.1074/jbc.M109.019570. Korber, M., Klein, I., Daum, G., 2017. Steryl ester synthesis, storage and hydrolysis: a Chi, Z., Arneborg, N., 1999. Relationship between lipid composition, frequency of contribution to sterol homeostasis. Biochim. Biophys. Acta 1862, 1534–1545. ethanol-induced respiratory deficient mutants, and ethanol tolerance in https://doi.org/10.1016/j.bbalip.2017.09.002. Saccharomyces cerevisiae. J. Appl. Microbiol. 86, 1047–1052. https://doi.org/10. Los, D.A., Murata, N., 2004. Membrane fluidity and its roles in the perception ofen- 1046/j.1365-2672.1999.00793.x. vironmental signals. Biochim. Biophys. Acta 1666, 142–157. https://doi.org/10. Cipak, A., Hasslacher, M., Tehlivets, O., Collinson, E.J., Zivkovic, M., Matijevic, T., 1016/j.bbamem.2004.08.002. Wonisch, W., Waeg, G., Dawes, I.W., Zarkovic, N., Kohlwein, S.D., 2006. Lowry, H., Nira, J., Rosebrough, A., Farr, Lewis, Randall, J., 1951. Protein measurement Saccharomyces cerevisiae strain expressing a plant fatty acid desaturase produces with the Folin phenol reagent. Anal. Biochem. 217, 220–230. https://doi.org/10. polyunsaturated fatty acids and is susceptible to oxidative stress induced by lipid 1016/0304-3894(92)87011-4. peroxidation. Free Radic. Biol. Med. 40, 897–906. https://doi.org/10.1016/j. Mas, A., Guillamón, J.M., Beltran, G., 2016. Editorial: non-conventional yeast in the wine freeradbiomed.2005.10.039. industry. Front. Microbiol. 7, 1494. https://doi.org/10.3389/fmicb.2016.01494. Costa, V., Moradas-Ferreira, P., 2001. Oxidative stress and signal transduction in Matallana, E., Aranda, A., 2017. Biotechnological impact of stress response on wine yeast. Saccharomyces cerevisiae: insights into ageing, apoptosis and diseases. Mol. Aspect. Lett. Appl. Microbiol. 64, 103–110 doi:10.111/lam.12677. Med. 22, 217–246. https://doi.org/10.1016/S0098-2997(01)00012-7. Moradas-Ferreira, P., Costa, V., 2000. Adaptive response of the yeast Saccharomyces Daum, G., Lees, N.D., Bard, M., Dickson, R., 1998. Biochemistry, cell biology and mole- cerevisiae to reactive oxygen species: defences, damage and death. Redox Rep. 5, cular biology of lipids of Saccharomyces cerevisiae. Yeast 14, 1471–1510. https://doi. 277–285. https://doi.org/10.1179/135100000101535816. org/10.1002/(SICI)1097-0061(199812)14:16<1471::AID-YEA353>3.0.CO;2-Y. Moradas-Ferreira, P., Costa, V., Piper, P., Mager, W., 1996. The molecular defences De Kroon, A.I., Rijken, P.J., De Smet, C.H., 2013. Checks and balances in membrane against reactive oxygen species in yeast. Mol. Microbiol. 19, 651–658. https://doi. phospholipid class and acyl chain homeostasis, the yeast perspective. Prog. Lipid Res. org/10.1046/j.1365-2958.1996.403940.x. 52, 374–394. https://doi.org/10.1016/j.plipres.2013.04.006. Murakami, Y., Yokoigawa, K., Kawai, F., Kawai, H., 1996. Lipid composition of com- Ding, J., Huang, X., Zhang, L., Zhao, N., Yang, D., Zhang, K., 2009. Tolerance and stress mercial bakers' yeasts having different freeze-tolerance in frozen dough. Biosci. response to ethanol in the yeast Saccharomyces cerevisiae. Appl. Microbiol. Biotechnol. Biotechnol. Biochem. 60, 1874–1876. https://doi.org/10.1271/bbb.60.1874. 85, 253. https://doi.org/10.1007/s00253-009-2223-1. Padilla, B., García-Fernández, D., González, B., Izidoro, I., Esteve-Zarzoso, B., Beltran, G., Fleet, G.H., Heard, G.M., 1993. Yeasts-growth during fermentation. In: Fleet, G.H. (Ed.), Mas, A., 2016. Yeast biodiversity from DOQ priorat uninoculated fermentations. Wine Microbiology and Biotechnology. Harwood Academic Publishers, Switzerland, Front. Microbiol. 7. https://doi.org/10.3389/fmicb.2016.00930. pp. 27–54. Padilla, B., Zulian, L., Ferreres, À., Pastor, R., Esteve-Zarzoso, B., Beltran, G., Mas, A., Fleet, G.H., 2008. Wine yeasts for the future. FEMS Yeast Res. 8, 979–995. https://doi. 2017. Sequential inoculation of native non-Saccharomyces and Saccharomyces cere- org/10.1111/j.1567-1364.2008.00427.x. visiae strains for wine making. Front. Microbiol. 8, 1–12. https://doi.org/10.3389/ Flis, V.V., Fankl, A., Ramprecht, C., Zellnig, G., Leitner, E., Hermetter, A., Daum, G., 2015. fmicb.2017.01293. Phosphatidylcholine supply to peroxisomes of the yeast Saccharomyces cerevisiae. Parks, L.W., 1978. Metabolism of sterols in yeast. Crit. Rev. Microbiol. 6, 301–340. PLoS One 10, 1–19. https://doi.org/10.1371/journal.pone.0135084. Pedroso, N., Matias, A.C., Cyrne, L., Antunes, F., Borges, C., Malhó, R., Rodrigo, F.M., Folch, J., Lees, M., Sloane, G.H., 1957. A simple method for the isolation and purification Herrero, E., Marinho, H., 2009. Modulation of plasma membrane lipid profile and

153 J. Vázquez et al. Food Microbiology 78 (2019) 143–154

microdomains by H2O2 in Saccharomyces cerevisiae. Free Radic. Biol. Med. 46, Schneiter, R., Brugger, B., Sandhoff, R., Zellnig, G., Leber, A., Lampl, M., Athenstaedt, K., 289–298. https://doi.org/10.1016/j.freeradbiomed.2008.10.039. Hrastnik, C., Eder, S., Daum, G., Paltauf, F., Wieland, F.T., Kohlwein, S.D., 1999. Pretorius, I.S., 2000. Tailoring wine yeast for the new millennium: novel approaches to Electrospray ionization tandem mass spectrometry (ESI-MS/MS) analysis of the lipid the ancient art of winemaking. Yeast 16, 675–729. https://doi.org/10.1002/1097- molecular species composition of yeast subcellular membranes reveals acyl chain- 0061(20000615)16:8<675::AID-YEA585>3.0.CO;2-B. based sorting/remodeling of distinct molecular species en route to the plasma Quail, M.A., Kelly, S.L., 1996. In: Evans, I.H. (Ed.), The Extraction and Analysis of Sterols membrane. J. Cell Biol. 146, 741–754. from Yeast BT - Yeast Protocols: Methods in Cell and Molecular Biology. Humana Spanova, M., Czabany, T., Zelling, G., Leitner, E., Hapala, I., Daum, G., 2010. Effect of Press, Totowa, NJ, pp. 123–131. https://doi.org/10.1385/0-89603-319-8:123. lipid particle biogenesis on the subcellular distribution of squalene in the yeast Querol, A., Fernández-Espinar, M.T., Del Olmo, M., Barrio, E., 2003. Adaptive evolution Saccharomyces cerevisiae. J. Biol. Chem. 285, 6127–6133. https://doi.org/10.1074/ of wine yeast. Int. J. Food Microbiol. 86, 3–10. https://doi.org/10.1016/S0168- jbc.M109.074229. 1605(03)00244-7. Spanova, M., Zweytick, D., Lohner, K., Klug, L., Leitner, E., Hermetter, A., Daum, G., Ratledge, C., Evans, C.T., 1989. Lipids and their metabolism. In: In: Rose, A.H., Harrison, 2012. Influence of squalene on lipid particle/droplet and membrane organization in J.S. (Eds.), The Yeasts, vol. 3. Academic Press, London, pp. 367–455. the yeast Saccharomyces cerevisiae. Biochim. Biophys. Acta Mol. Cell Biol. Lipids 1821, Redón, M., Guillamón, J.M., Mas, A., Rozès, N., 2009. Effect of lipid supplementation 647–653. https://doi.org/10.1016/j.bbalip.2012.01.015. upon Saccharomyces cerevisiae lipid composition and fermentation performance at Steels, E.L., Learmonth, R.P., Watson, K., 1994. Stress tolerance and membrane lipid low temperature. Eur. Food Res. Technol. 228, 833–840. https://doi.org/10.1007/ unsaturation in Saccharomyces cerevisiae grown aerobically or anaerobically. s00217-008-0996-6. Microbiology 140 (Pt 3), 569–576. https://doi.org/10.1099/00221287-140-3-569. Rhoads, D.M., Umbach, A.L., Subbaiah, C.C., Siedow, J.N., 2006. Mitochondrial reactive Stukey, J.E., McDonough, V.M., Martin, C.E., 1990. The OLE1 gene of Saccharomyces oxygen species. Contribution to oxidative stress and innerorganellar signaling. Plant cerevisiae encodes the delta 9 fatty acid desaturase and can be functionally replaced Physiol. 141, 357–366. https://doi.org/10.1104/pp.106.079129. by the rat stearoyl-CoA desaturase gene. J. Biol. Chem. 265, 20144–20149. Ribereau-Gayon, P., 1985. New developments in wine microbiology. Am. J. Enol. Vitic. Suutari, M., Laakso, S., 1994. Microbial fatty acids and thermal adaptation. Crit. Rev. 36, 1–10. Microbiol. 20, 285–328. https://doi.org/10.3109/10408419409113560. Rodríguez-Vargas, S., Sánchez-García, A., Martínez-Rivas, J.M., Prieto, J.A., Randez-Gil, Torija, M.J., Beltran, G., Novo, M., Poblet, M., Guillamón, J.M., Mas, A., Rozès, N., 2003. F., 2007. Fluidization of membrane lipids enhances the tolerance of Saccharomyces Effects of fermentation temperature and Saccharomyces species on the cell fatty acid cerevisiae to freezing and salt stress. Appl. Environ. Microbiol. 73, 110–116. https:// composition and presence of volatile compounds in wine. Int. J. Food Microbiol. 85, doi.org/10.1128/AEM.01360-06. 127–136. https://doi.org/10.1016/S0168-1605(02)00506-8. Romano, P., Fiore, C., Paraggio, M., Caruso, M., Capece, A., 2003. Function of yeast Vázquez, J., González, B., Sempere, V., Mas, A., Torija, M.J., Beltran, G., 2017. Melatonin species and strains in wine flavour. Int. J. Food Microbiol. 86, 169–180. https://doi. reduces oxidative stress damage induced by hydrogen peroxide in Saccharomyces org/10.1016/S0168-1605(03)00290-3. cerevisiae. Front. Microbiol. 8, 1–14. https://doi.org/10.3389/fmicb.2017.01066. Rosi, I., Bertuccioli, M., 1992. Influences of lipid addition on fatty acid composition of Vendramin-Pintar, M., Jernejc, K., Cimerman, A., 1995. A comparative study op lipid Saccharomyces cerevisiae and aroma characteristics of experimental wines 1. J. Inst. composition of baker's and brewer's yeasts. Food Biotechnol. 9, 207–215. https://doi. Brew. 98, 305–314. https://doi.org/10.1002/j.2050-0416.1992.tb01113.x. org/10.1080/08905439509549894. Rossouw, D., Heyns, E.H., Setati, M.E., Bosch, S., Bauer, F.F., 2013. Adjustment of tre- Voelker, D.R., Frazier, J.L., 1986. Isolation and characterization of a Chinese hamster halose metabolism in wine Saccharomyces cerevisiae strains to modify ethanol yields. ovary cell line requiring ethanolamine or phosphatidylserine for growth and ex- Appl. Environ. Microbiol. 79, 5197–5207. https://doi.org/10.1128/AEM.00964-13. hibiting defective phosphatidylserine synthase activity. J. Biol. Chem. 261, Ruenwai, R., Neiss, A., Laoteng, K., Vongsangnak, W., Dalfard, A.B., Cheevadhanarak, S., 1002–1008. Petranovic, D., Nielsen, J., 2011. Heterologous production of polyunsaturated fatty Walker, G., Dijck, P., 2006. Physiological and molecular responses of yeasts to the en- acids in Saccharomyces cerevisiae causes a global transcriptional response resulting in vironment. In: Querol, A., Fleet, G.H. (Eds.), Yeasts in Food and Beverages. Springer- reduced proteasomal activity and increased oxidative stress. Biotechnol. J. 6, Verlag, Berlin Heidelberg, pp. 111–112. 343–356. https://doi.org/10.1002/biot.201000316. Xia, J., Jones, A.D., Lau, M.W., Yuan, Y.J., Dale, B.E., Balan, V., 2011. Comparative li- Rußmayer, H., Buchetics, M., Gruber, C., Valli, M., Grillitsch, K., Modarres, G., Guerrasio, pidomic profiling of xylose-metabolizing S. cerevisiae and its parental strain indif- R., Klavins, K., Neubauer, S., Drexler, H., Steiger, M., Troyer, C., Al Chalabi, A., ferent media reveals correlations between membrane lipids and fermentation capa- Krebiehl, G., Sonntag, D., Zellnig, G., Daum, G., Graf, A.B., Altmann, F., city. Biotechnol. Bioeng. 108, 12–21. https://doi.org/10.1002/bit.22910. Koellensperger, G., Hann, S., Sauer, M., Mattanovich, D., Gasser, B., 2015. Systems- Yazawa, H., Iwahashi, H., Kamisaka, Y., Kimura, K., Uemura, H., 2009. Production of level organization of yeast methylotrophic lifestyle. BMC Biol. 13, 80. https://doi. polyunsaturated fatty acids in yeast Saccharomyces cerevisiae and its relation to al- org/10.1186/s12915-015-0186-5. kaline pH tolerance. Yeast 26, 167–184. https://doi.org/10.1002/yea.1659. Santomartino, R., Riego-Ruiz, L., Bianchi, M.M., 2017. Three, two, one yeast fatty acid You, K.M., Rosenfield, C., Knipple, D.C., 2003. Ethanol tolerance in theyeast desaturases: regulation and function. World J. Microbiol. Biotechnol. 33, 1–12. Saccharomyces cerevisiae is dependent on cellular oleic acid content. Appl. Environ. https://doi.org/10.1007/s11274-017-2257-y. Microbiol. 69, 1499. https://doi.org/10.1128/AEM.69.3.1499.

154